NOTES AND DISCUSSION

Reply to Fairclough: Discourse and interpretation:

conjectures and refutations
1
H.G. Widdowson, University of London, University of Essex, UK
Norman Fairclough and I disagree about a lot of things. That much is obvious.
But let me begin my comments by acknowledging the importance of his
contribution to our thinking about language and society. He has brought to our
attention a whole range of issues about the dialectical interplay of language and
social life, not as matters for detached academic debate but as immediately
implicated in practical and political affairs. He has alerted us to how language
can be exploited in the manipulation of opinion and the abuse of power. His
work is impressive in scope and purpose. It is highly stimulating for those who
share his views, highly provocative for those who do not, and is to be valued on
both counts.
It is also highly influential. Indeed it can be said to have been the main force
in the establishing of the new paradigm (or episteme, to use Foucault’s term?) of
Critical Discourse Analysis. This can be seen as a new ideological orthodoxy
and as such, paradoxically, it exerts just the kind of discursive domination which
it seeks to expose in other uses of language. It is because the work of Fairclough
and his colleagues has such significance as a line of enquiry, and because it has
been so influential as a discursive practice in its own right, that it invites critical
attention. This is why I wrote the article that Fairclough comments on. I wanted
to point to what seemed to me to be fundamental problems in their approach to
analysis.
His indicates that all I succeeded in doing was to misrepresent him. This
reply
is not what I
intended. There was no deliberate distortion, but only my
interpretation of what he had written. But then how could I have got it so
wrong? How could I have read such mistaken meanings into his text? One
reason, as Fairclough’s comments make abundantly clear, is that I do not share
the same ideological position on the nature of discourse. I do not see things his
way. So I read his text in ways other than he intended, and in ways other than
those discursively aligned with him. Just so. ,That is precisely the point I was
trying to make. k

And by the same token, my own text is subject to misrepresentation. One

example. I am charged with being too liberal in supposing that the ’social’ is ’a
voluntary association of free individuals’, that I disregard social conditions and
see the pragmatic achievement of meaning as ’simply a matter of circumstances
or will of the individual, graced as a &dquo;natural&dquo; pragmatic process.’ I should be
interested know where there is textual warrant in my work for this bizarre
to
interpretation of my position. You cannot study discourse in disregard of social
factors and I do not know of anybody who claims you can. But the study of
discourse is not just the study of how it is socially constituted but of how it is

Language and Literature 1996 5 (1) 57-69 © Addison Wesley Longman Ltd 1996 0963-9470/96/05106057/$03.50
58

pragmatically realised, not only by conforming to its constitution, but also by

subverting it.
Norman Fairclough complains that discourse analysis, on my account, ’is
reduced to pragmatics’. Why reduced? Discourse, in my conception of it, cannot
be reduced to pragmatics. It is, for me, crucially, a function of pragmatics: the
process whereby different interpretations are drawn from the textual data. Of
course this process implicates all manner of social factors: assumptions, beliefs,
values, ideologies, which would fall within a Foucault concept of discourse. So
it would seem that we need to distinguish two senses of discourse (see
Widdowson 1990). The discourse process in this pragmatic sense (Discourse 1,
we might call it) is influenced by the different discourses that participants have
been socialised in (Discourse 2). But it is not determined by them. To suppose
that it is is indeed to reduce discourse. And this is probably where my position
differs most radically from that of Norman Fairclough and his colleagues. So let
me make it as explicit as I can.
I do not believe that individuals simply act out social roles. There are socially
constituted Discourses 2 - conventions of belief, established values which
constrain the way people think and use their language to achieve meaning. But
people’s activities are not determined by their ideological allegiances. They are
not bound by them. You can of course ascribe social roles to individuals and part
of their individuality can obviously be associated with this group identity.
Pragmatics would take account of this in its consideration of contextual
conditions. But to think of individuals as if they were representative of such
groups, as tokens of the type, is to deal in stereotypical constructs, well defined
social categories. And these are abstractions from reality analagous to those of
formalist linguistics: ’discursively constructed social subjects’ seem to me to be
beings of the same order of abstraction as ideal speaker-listeners in homogenous
speech communities.
For me, discourse is individual engagement. It is individual, not social
subjects, who interact with each other. Of course, I do not mean to suggest that
they are free agents to do what they will. They are constrained by established
conventions and regulations, and restrictions are set on their initiative. But they
are not absolutely controlled by them: there is always room for manoevre. That
is our salvation. Whatever communal ideological values are institutionally in
power, they can be, and constantly are, subverted by individual initiative, for
better or for worse. So it is that suppressed &dquo;ideas go underground and bide their
time. So it is, too, that in societies dedicated to equality, individuals are busy
feathering their own nest, and in those dedicated to morality, the custodians of
public probity indulge-their vices in private. Individuals are never
institutionalised: they retain their contradictory idiosycracies, their impulses,
fears, sensitivities and prejudices. And when they find others who share some of
these, they give them institutional status, call them principles, form a group,
draw up a constitution, found a party, a sect, a school.
But the very fact that individuals can bring about social change makes it clear
 59

that social forces/even the malevolent imposition of power, can have positive
consequences in spite of themselves. This, I believe, is because the very
constraints which inhibit freedom of thought also define the conditions which
enable it to be exercised. And this is as true of language as it is of society.
Language sets limits on what people can say, but those limits also allow them to
exploit possibilities of meaning immanent in the language but not conventionally
encoded or customarily used. Society exacts a certain conformity in the same
way. But the very values and beliefs which constitute it can be exploited to
realign them. Thus Norman Fairclough, as an individual thinker of considerable
originality, employs an established ’mode of argumentation and mode of
rationality’ (as he puts it) in order to criticise the social values on which it is
based. It seems to me to be entirely reasonable that he should. Indeed, he has no
choice. No change, no matter how apparently revolutionary, was ever entirely
innovative. Language and social change are a matter of reformulation not
invention: the essential features remain in place, but differently aligned,
differently expounded. This indeed is implied by Fairclough’s own concept of
intertextuality, whereby different modes of argumentation and rationality are
created by just this kind of realignment. I am pleased to find that this concept
(unless I have misrepresented it) corresponds quite closely to my own thinking
(see Widdowson 1988). Happily there are some things that Norman and I can
agree about.
So I certainly do not see society as ’a voluntary association of free
individuals’. Of course the individuals are constrained, but they remain
. individuals none the less: they are not just ’subjects discoursally constituted’.
This is not to deny the existence of discourses in the Foucault sense as
conventionalised modes of knowledge (i.e. Discourse 2), nor the importance of
studying the discursive construction of social subjects at an appropriate level of
idealisation. But these discourses are abstract constructs. They can only be
actualised through discourse as I have defined it, as the pragmatic process of
meaning negotiation (i.e. Discourse 1). The difficulty is knowing what features
of the resulting text can be adduced as evidence of the discursively constructed
subject. This is where the problematic issue of interpretation comes in, of which
more presently.
My position, then, is that different discourses in the Foucault sense, as
ideological and idealised social constructs can only be activated through
discourse in my sense: through the pragmatics of individual interaction. So it is
that this present exchange between Norman Fairclough and myself is not
between political values or ideological positions, Marxist versus Liberal or
whatever, but between him and me, his views, sensitivities, prejudices, and mine.
These are, of course, formed in part by our social history and political
allegiances, expressive to some degree of the different discourse communities we
belong to, but essentially this discourse we are pragmatically engaged in is a
personal matter. In his comments Norman Fairclough refers to his own work
rather than CDA as a whole, I would guess that it was his resentment of what he
60

saw as my misrepresentation of his work which provoked his response. And that,
in my view, is as it should be.
How then have I misrepresented his work? A key problem would seem to be
that I have failed to distinguish between two senses of the term ’interpretation’.
Fairclough claims that explanation (Interpretation 2) is a matter of showing how
the response to a text (Interpretation 1) is influenced (perhaps determined) by
social factors - how this interpretation is, in effect, socially constructed. So
Interpretation 2 is an interpretation of Interpretation 1. But then we surely need
to establish first what Interpretation 1 is by consulting those who practice it.
Fairclough does not see it this way. His belief is that he can gain access to
Interpretation 1 through introspection. He comments:
There is a similarity here between my account in Chapter 3 of what the

analyst does, and what text interpreters do: analysts too need the resources
they have as competent members of communities, even if they use them
rather more systematically.’
(Fairclough 1992: 198)
This position does not seem to me to be essentially different from that taken by
formal linguists: as representative members of the speech community whose
language they describe, they can serve as the source of their own data and do not
need to consult other informants. But, as has been pointed out often enough, they
are not reliable as informants themselves precisely because they are analysts.
And the idea that analysts can be competent members of the communities whose
language use they analyse would seem to be in complete contradiction with
Fairclough’s own theoretical position, which is that there are different
communities with distinct and competing discourses. People (whether ’ordinary
text interpreters’ or analysts) cannot be competent members of all of them. If
they could, there would presumably be no covert domination, no hegemonic
struggle, and indeed nothing for Critical Discourse Analysis to be critical about.
It seems to me that if the analyst provides Interpretation 1 by proxy, it is
bound to get confused with Interpretation 2, and that, I think, is precisely what
happens. Fairclough says that he does not just read off significance from the
textual facts themselves, and has taken other critical discourse analysts to task
for so doing. But it is hard to see how he can do otherwise if he relies on his
own intuitions, since he cannot be a competent member of all the discourse
communities whose texts he analyses. He recognises that there may, in principle,
be a plurality of interpretations, and so a plurality of effects. Presumably, then,
we should expect from his analysis that it would provide an explanation of what

different social factors motivate this plurality. But it does not.

I want now to illustrate these difficulties in reference to a specific example of
analysis. In so doing I shall subject Fairclough’s own text to the same kind of
close critical scrutiny that he applies to others, but without drawing conclusions
about its ideological significance.
 61

Discourse and Social Change (Fairclough 1992) develops a theory of

discourse as hegemonic struggle whereby power is exercised to construct social
reality by the intertextual control of discursive practices. The second part of the
book provides analysis to support the theory expounded in the first part. The
example I shall consider is from Chapter 6, which is entitled: ’Text Analysis:
Constructing Social Reality’. The first analysis is of a text sample taken from a
booklet about pregnancy. The sample is as follows:

Antenatal care
The essential aim of antenantal care is to ensure that you go through

pregnancy and labour in the peak of condition. Inevitably, therefore, it

involves a series of examinations and tests throughout the course of your
pregnancy. As mentioned above, antenatal care will be provided either by
your local hospital or by your general practitioner, frequently working in
cooperation with the hospital.
It is important to attent for your first examination as early as possible, since
there may be minor disorders that the doctor can correct which will benefit
the rest of your pregnancy. More particularly, having seen your doctor and
booked in at a local hospital, you will usually receive the assurance that
everything is proceeding normally.
The first visit
Your first visit involves a comprehensive review of your health through
childhood and also right up to the time you became pregnant. Just
occassionally [sic] women may suffer from certain medical disorders of
which they are unaware - such as high blood pressure, diabetes and kidney
disease. It is important for these problems to be identified at an early stage
since they may seriously influence the course of the pregnancy.
The doctor and the midwife will also want to know about all your previous
health problems, as well as discussing your social circumstances. We do know
that social conditions can influence the outcome of the pregnancy. For this -

reason, they will ask you details about your housing, as well as your present
job. In addition they will need to know if you smoke, drink alcohol or if you
are taking any drugs which have been prescribed by your doctor or chemists.
All of these substances can sometimes affect the development of a baby.
.,

Examination
You will be weighed so that your subsequent weight gain can be assessed.
Your height will be-measured, since small women on the whole have a
slightly smaller pelvis than tall women - which is not surprising. A complete
physical examination will then be carried out which will include checking
your breasts, heart, lungs, blood pressure, abdomen and pelvis.
The purpose of this is to identify any abnormalities which might be present,
but which so far have not caused you any problems. A vaginal examination
62

will enable the pelvis to be assessed in order to check the condition of the
uterus, cervix and the vagina. A cervical smear is also often taken at this time
to exclude any early pre-cancerous change which rarely may be present.
(Quoted in Fairclough 1992: 170-1; originally
from N. Morris (ed.) The Baby Book.)

The first thing to notice is that what we have here is a text fragment. We do not
know how it functions in relation to the rest of the booklet. Furthermore, the
three extracts that we have are discontinuous: a whole subsection between the
second and third has been omitted. No reason is given for this. We do not know
either anything about what the motivation for the text was: if, or why, it was
commissioned, and by whom. There is good deal of talk in Discourse and Social
Change (and indeed in Fairclough’s work generally) about the production and
consumption of texts, but in this case, and in the case of other analyses, the
actual producers and consumers are not consulted as to what their intentions or
be.
interpretations might Instead, these are ascribed on the sole evidence of the

analysis: in other words, they are read into the text. But not only is the sample a
fragment, the is
analysis fragmented too, for it starts not, as one might have
expected, from the first section, but from the last. Whatever reason there might
be for this is not explained. We can only assume that it is random.
The analysis of this section reveals that there is a predominance of sentences
which consist of two clauses (referred to oddly enough as simple sentences)
expounding purpose and reason relations:
Clause 1 so thai since in order to Clause 2
(Fairclough 1992: 171)

These are textual facts which nobody I imagine would wish to dispute. But this
it what Fairclough refers to as the description dimension. The facts have now to
be interpreted. There is the difficulty, though, that description itself to some
degree implies interpretation. Fairclough himself points this out, but compounds
the difficulty by ignoring the two senses of the term he has earlier distinguished,
so we do not know whether he is talking about Interpretation 1 or 2, or both:

Description is not as separate from interpretation [2?] as it is often assumed

to be. As an analyst (and as an ordinary text interpreter) one is inevitably

interpreting [1? 2?] all the time, and there is no phrase of analysis which is
pure description. Consequently, one’s analysis of the text is ’shaped and
coloured’ by one’s interpretation [2?] of its relationship to discourse processes
’

and wider social processes.

(Fairclough 1992: 199, my additions in
square brackets).

But here, perhaps we have an indication as to why these particular fragments

 63

were extracted from the whole text, and why the analysis starts at apparent
random: these procedures are shaped and coloured in advance by the discourse
processes and wider social processes that they are designed to discover. The
analyst preconceives the data as evidence. Fairclough knows what he is looking
for. Not suprisingly, he finds it.
But what is the interpretation of this particular description of clauses?

The message that comes across is one of re-assurance: everything that

happens during antenatal care is there for a good reason.
(Fairclough 1992: 171-2)
How does Fairclough know that the readers for whom this text was designed
(the consumers) are reassured by it? This conclusion does not emerge from the
analysis, but is imposed upon it. One might reasonably conclude from internal
textual evidence that the intended illocutionary force of this passage is
explanation. But reassurance is a perlocutionary effect. The only way of finding
out whether this is, indeed, the effect is to ask the pregnant women for whom it
is written. One might even consider asking the producers whether this was
intended. But you can only read such an effect off from the text like this by
assuming a vicarious identity. This I would suggest is a case of Interpretation I
masquerading as Interpretation 2.
But the analyst here not only assumes the identity of the consumer but of the
producer as well. The description which follows seeks to demonstrate that the
dominant discourse is a medico-scientific one, and the social reality of
pregnancy is constructed in its terms. The pregnant women are positioned as
compliant patients (grammatically and medically). The medical staff are in
control. Evidence for this is found in the second section of the extract where
there is a shift from third to first person (’The doctor and midwife’ to ’We’).
This apparently makes it clear that the text is produced by medical staff: it is
their voice which speaks. In the next piece of text, however, we are told there is
evidence of two voices in the following sentence:

Your height will be measured, since small women on the whole have slightly
smaller pelvis than tall women - which is not surprising.
‘
Fairclough’s interpretation runs as follows:
... the tagged-on comment ’which is not surprising’ comes across as the
lifeworld voice of the-prospective patient, or indeed of the medical staff in
’

their non-professional capacities ...

(Fairclough 1992: 172)

Again, we might ask ’comes across to whom?’ We just do not know whether this
comes across to the prospective patients since they have not been consulted. But
64

anyway, if this voice is indistinguishable from that of the medical staff ’in their
non-professional capacities’ how can we talk about two voices anyway? Are we
to suppose that the medical voice is always uniform and unmodulated by
lifeworld concerns, that medical staff in their professional capacities are
necessarily and inevitably detached, objective, and technical in their treatment of
patients? The analysis continues:
But notice the contrast in voices between this and the second clause of the
sentence (’since small women on the whole have a slightly smaller pelvis than
tall women’), which is a reason clause. This is in the medical voice: ’pelvis’
is a medical term, the clause consists of an authoritative assertion, which we
take to be grounded in medico-scientific evidence. It is also far more typical
of the extract as a whole: most reason clauses are in the medical voice.

What is it that characterises this medical voice? What textual evidence can we
adduce to identify it? Technical vocabulary perhaps: ’pelvis’ is a medical term. Is
it? So what, one might ask is the non-medical equivalent? And what of ’breasts’,
’heart’, ’lungs’, ’blood pressure’, ’abdomen’? Are these all medical terms as
well? The point is, surely, that you cannot talk about pregnancy at all, in any
voice, without using terms like this. So if the patients were to employ them,
would they too be using contrasting voices, enacting some hegemonic discursive
struggle by using non-lifeworld capacities?
But if it is not lexis which is the defining feature, then what is? We are told
that ‘most reason clauses are in the medical voice’. Some are, then, and some are
not. Which is which? We are not told, but unless we know what makes a reason
clause medical as distinct from lifeworld, there is no way of knowing that the
particular clause we are considering (’since small women on the whole have a
slightly smaller pelvis than tall women’) is ’f~r more typical of the extract as a
whole’ than the phrase ’which is not surprising’. What is it that makes it typical?
We have no way of knowing because the type is left unspecified. We are told
that ’this clause consists of an authoritative assertion, which we take to be
grounded in medico-scientific evidence’. But the clause consists of nothing of
the kind. It consists of linguistic constituents. Whether it is an assertion or not,
and certainly whether it is an authoritative assertion or not, and how it is taken
to be grounded, are interpretations based on an assumption of attitude assigned
to the producer in advance. ’... whicnwe take as grounded ...’ Who is ’we’?
Do the prospective patients take it in this way? This is Fairclough speaking on
behalf of other readers. The explanation is an instance of Interpretation 1, but
not, notice, an Interpretation 1 of the people for whom the text was written, but
one vicariously assigned to them by the analyst.

But we have not yet finished with this particular clause. The phrase ’on the
whole’ calls for comment. This is what Fairclough says:
 65

The hedging of the assertion (’on the whole’) is interesting. On the one hand
its vagueness suggests a shift into the voice of the lifeworld, while on the
other it marks the cautious and circumspect ethos we associate with scientific
medicine.
(Fairclough 1992: 173)
So in this case there is no way of knowing which voice is operative. But even if
we opted for the lifeworld interpretation, it could still be taken as the voice of
medico-scientific authority, but subtly modulated so that it sounds as if it was
not. Fairclough goes on to provide further evidence of this ruse. First he
summarises the findings of his analysis of the given extract:

The clauses of reason or purpose, consistently cast in the voice of medicine,

give the sort of rationalization and argumentation one would expect from
medical staff, which contributes to the construction of medico-scientific ethos
in the extract.
(Fairclough 1992: 173)
The clauses of reason and purpose, we should note, are now represented as
coizsistetitly expressive of the medical voice: not most, but all of them, are cast
in this idiom. They contribute to ’the construction of a medico-scientific ethos’.
What else contributes we are not told. But we are given an example of another
text, one which contrasts with it, taken this time from a publication called The
Pregnancy Book. This reads as follows:
’

Throughout your pregnancy you will have regular check-ups ... This is to niake
siti-e that both yozi arrd the baby are fit and well, to check that tlte baby is

developing properly, and, as far as possible to prewfnt anything going wrong...

(Quoted in Fairclough 1992: 173)
The italics are provided by Fairclough. He comments:

The italicized expressions are evidently closer to the voice of the lifeworld
than equivalent ones in The Baby Book ...

-,
(Fairclough 1992: 173)
If these are evidently closer, what is the evidence? Why, one is bound to wonder,
is ’check-up’ not italicised as lifeworld (cf. the equivalent terms ’examination’
, and ’test’, which figure in the medico-scientific Baby Book). Why, on the other
hand are ’This is’ and ’and’ not italicised? How can the single use of ’and’ mark
a shift into a different voice? So in what respect, exactly, are these italicised

expressions ’closer to the voice of the lifeworld’? To the extent that they
comprise clauses of reason and purpose one might suppose, on Fairclough’s own
argument, they are just as close to the medical voice. Perhaps this is why
66

Fairclough himself seems uncertain about their status. He continues:

... but I feel nevertheless that there is an ambivalence of voice in the

Pi-egizaticy Book.
(Fairclough 1992: 173)

We might expect that reference might now be made to these clauses, or to some
other textual features to lend support to this feeling. But we get nothing of the
kind:

The reason for this is that medical staff often do shift partly into a lifeworld
voice when talking to patients ... and the italicized expressions could be used
by medical staff. It therefore remains unclear whether the producer of the
Pregnanc)’ Book is writing from the patient’s perspective, or from that of (a
’modernizing’ position among) medical staff.
(Fairclough 1992: 173)
But this is an interpretation without description. It is simply an assertion,
unfounded in this particular text. It remains unclear whether the text producer is
using one voice or another because there are no clear textual criteria which
would enable us to distinguish them. We would not know whether there is a shift
or not, and even if we had convincing linguistic reasons to suppose there was,

we would not know whether this textual shift implied a corresponding discourse

shift, or was simply a rhetorical ploy.

The medical voice is elusive, and yet it is crucial to Fairclough’s argumentt
that it should be identified since, otherwise he cannot demonstrate intertextuality,
and show how the discursive practices of authority position people by borrowing
other voices, thereby exercising deception by disguise. So medical people
pretend to adopt lifeworld values in much the same way as advertisers assume
the guise of advisers and politicians adopt the idiom of ordinary talk.
Fairclough’s contention, and it is central to his theory of discourse and social
change, is that people in power shift voices in subtle ways to exert their
influence. In the present case, medical staff will assume a voice expressive of a
patient perspective so as to keep them in their place.
The whole argument depends on the assumption that there are separate and
distinguishable discourses. There is a community of social subjects, the medical
staff, which has its own distinct discourse expressive of its own distinct ethos.
The position it adopts necessarily precludes the perspective of the patient, which
is a different social subject, belonging to a different discourse - a lifeworld one.
So although medical people may appear to adopt this perspective they do so only
as a tactical ruse to win over the patient by a kind of covert intertextual
colonisation.
You can, of course, read this significance into these texts if you are willed so
to do. But this interpretation is supposed to be to some extent at least dependent
 67

on a description of the text. In the analysis I have considered here, it seems that
the reverseis the case: description is provided to support a preconceived
interpretation. And this interpretation is Interpretation 1. It cannot be
Interpretation 2, an explanation of how a particular reading is influenced by
social forces, since we have no data whatever on what significance the intended
consumers do read into these texts. We only have the analyst taking over their
role. Paradoxically, therefore, we find Fairclough indulging in the very
discursive practice he sets out to expose: assuming the voice of the patient for
the purpose of persuasion.
This is not to say that the interpretation is without value or without textual
warrant, only that it is partial. Partiality is a necessary pragmatic feature of any
interpretation, whatever term you use for it, and in my view it is the main
business of analysis to explain how this partiality effects what you read into a
text. These comments I have offered on Fairclough’s findings are analysis in this
sense. But then analysis should also reveal the essential instability of textual

meaning by indicating how they can give rise to alternative interpretations. Do

the texts that FaircIough deals with admit of different readings? Let us, briefly,
consider the question.
These textual extracts are about pregnancy. This is both a physical process
that can be seen as a third person occurence, something that involves the human
body, and a personal experience, something that involves the individual human
being. So there are obviously two very different ways of talking about it:
objectively as fact, from the perspective of the outsider observer; and
subjectively as affect from the perspective of the insider participant. Medical
staff naturally take up the first position, and prospective mothers the second:
naturally, because these positions are determined by the nature of the process not
by any socially sanctioned assignment of role. But the purpose of the interaction
between these parties is in some way to mediate between them. The prospective
mothers need to know something about their bodies, the medical staff on the
other hand can only talk to the human beings. Any medical text, therefore, is
likely to have features which reflect this dual perspective.
So it is that in our present text we find a continual shifting from
non-participant third person to participant second person reference which reflects
a relative distancing from the concerns of the individual. The pregnant women
are both addressed and talked about. Thus the participant phrases ’your

pregnancy’, ‘your local hospital’, ‘your general practitioner’ co-occur with the
non-participant equivalents ’the doctors local hospial’, ’the pregnancy’. But
the participant expressions always come first and the non-participant ones take
on a dependent anaphoric function. One way of interpreting these textual facts is
to suggest that the writer’s first concern is to acknowledge the insider
perspective.
Not all non-participant expressions are anaphoric however. In the third
extract, we find reference to both ‘your pelvis’ and ‘the pelvis’. But the former is
associated with ’your breasts, heart, lungs, blood pressure, abdomen’ and the
68

latter with ‘tlre uterus, cervix and the, vagina’. We might note that the former list
becomes lexically more medical as if marking a gradual distancing into
non-participant reference. This then becomes dominant in the second paragraph
where is only one single occurrence of the second person (’which so far have not.
caused you any problems’). In this paragraph, we might suggest, the medical
perspective takes over - appropriately enough, one might suggest, since this
particular extract is, by its title quite explicitly about examination. It is about
human bodies, not human beings.
But the medical perspective, we should note, comes in the second paragraph.
Again it is the participant perspective which comes first. This would account for
the occurence of the phrase ’which is not surprising’. For Fairclough, as we have
seen, this is the interpolation of a contrasting lifeworld voice, a kind of
’slippage’, like the occurence of ’we’ in the second extract. But they can also be
read as support for the participant orientation which is predominant in the
paragraphs they occur in. Actually the paragraph in whic’h ’we’ appears is the
second paragraph of the extract, not the first, and it might seem that the principle
of establishing participation before medical distancing is not evident in this case.
But notice that participation is already keyed into the beginning of this first
paragraph (’Your first visit’, ‘your health’, ‘you became pregnant’) and then in
the remaining two sentences we find a shift entirely into the non-participant
mode with no second person presence whatever.
What I hope these observations indicate is that there is a pattern discernible in
this text which Fairclough makes no mention of, and which gives rise to a rather
different interpretation of its purpose and possible effect: one which is more
co-operative than conflictual, which invokes no hegemonic struggle, and which
is rather more favourable to the medical profession. I would argue that
Fairclough does not notice this pattern because he is looking for something else.
And of course he might argue that I have only noticed it because I am
determined to be contrary. This is probably true. I would not have subjected the
text to scrutiny if he had not provoked me to do so. But this is just the point. We
came to the same text with different motives, assumptions, beliefs, values and so
read our different discourses into it.
I make no claims for this brief analysis. It, too, is bound to be partial, and I
have no idea how far my reading would be consistent with the effect it might
have on the women for whom it was written. But we could try to find out. One
way of doing this would be to elicit their’reactions to the original texts and to
versions of them in which the linguistic features I have identified have been
systematically altered. One version of the first extract might be:
The essential aim of antenatal care is to ensure that women go through
pregnancy and labour in the peak of condition. Inevitably, therefore it
involves a series of examinations and tests throughout the course of the
pregnancy. As mentioned above, antenantal care will be provided either by
the local hospital or by tie general practitioner ...
 69

The beginning of the second extract might be altered to read:

The first visit involves a comprehensive review of the patient’s health through
childhood and also right up to the time of the pregnancy...

And so on.
As Fairclough himself points out, no study of language is neutral. Idealisation
implies ideology in some degree: you cast reality in your own mental image.
This is bound to affect interpretation, and distinguishing two senses of the term
in principle does not help if the two are conflated in practice, as I believe they
are in the analysis we have considered. Fairclough’s concept of distinct voices,
of homogeneous groups with uniform discursive practices is well suited to his
image of the hegemonic struggle: social subject against social subject, doctor
against patient, man again woman, them against us.
And him against me? I think the discourse we are here enacting is a matter of
individuals arguing against each other. We take up different positions, and these
are of course to some degree discursively informed, expressive of different
modes of thought, political beliefs, social values. And so if we feel frustrated and
affronted, we might be tempted to reduce each other to ’social subjects
discursively determined and constructed’ and I cry ’Marxist!’ and he cries
’Liberal’ and we resentfully go our separate ways. But that would be a pity. Our
positions are not as polarised as all that. I am no apologist for ’greater
linguistics’ and totally reject the intertextual implication of the term. I agree that
we need to take an interdisciplinary perspective on language study. I agree that

people should be alerted to language abuse, and made aware of the ways in
which it can be used to persuasive and manipulative effect. You do not have to
subscribe to CDA to believe these things, and you do not have to be a critical
linguist to have a social conscience. So I think there is a good deal of common
ground between Norman Fairclough and myself, which both of us ought to be
willing to explore.

Note

1. My thanks to Barbara Seidlhofer for comments on an earlier draft of this article.

References

Fairclough, N. (1992) Discourse and Social Change, Polity Press, London

Widdowson, H. G. (1988) Language spread in modes of use, in P. H. Lowenberg (ed.) Language
Spread and Language Policy, Georgetown University Press, Georgetown USA
Widdowson, H. G. (1990) Discourses of enquiry and conditions of relevance, in J. Alatis (ed.)
Linguistics, Language Teaching and Language Acquisition: The Interdependence of Theory.
Practice and Research, Georgetown University Press, Georgetown, USA